Yesterday we received some data from Tick Data, Inc. (www.tickdata.com) and were impressed by how clean it was compared to what we had been getting from Quote.com. We asked the CEO of Tick Data why it was so clean and this was his reply:
“We converge multiple data feeds to ensure continuous data, then run through 3 scrubbing algorithms and 2 time scrubbing algorithms. If any interval greater than 3 minutes is missing, we go to additional datafeeds. I have 3 people who do nothing but scrub data full time.”
Therefore, if you want to build some historical models with very clean data, get some text files from Tick Data. Later you can run those models for trading with Quote.com. Tell them to send you text files compatible with NeuroShell.